corpus.ulaval.ca · Web view2020. 7. 29. · We followed the administration procedures and...
Transcript of corpus.ulaval.ca · Web view2020. 7. 29. · We followed the administration procedures and...
TMT: Normative Data for Quebec French-speaking Adults
Trail Making Test A and B: Regression-Based Normative Data
for Quebec French-speaking Mid and Older-Aged Adults
Alexandre St-Hilaire1, Camille Parent1,2, Olivier Potvin1, Louis Bherer3,4,5,
Jean-François Gagnon6,7, Sven Joubert4,8, Sylvie Belleville4,8, Maximiliano A.
Wilson1,9, Lisa Koski10,11, Isabelle Rouleau6,12, Carol Hudon1,2 & Joël Macoir1,9
1Centre de recherche CERVO, Institut universitaire en santé mentale de Québec, Québec, QC,
Canada2École de psychologie, Université Laval, Québec, QC, Canada3Department de médecine, Université de Montréal, Montréal, QC, Canada4Centre de recherche, Institut universitaire de gériatrie de Montréal, Montréal, QC, Canada5Centre de recherche, Institut de Cardiologie de Montréal, Montréal, QC, Canada6Département de psychologie, Université du Québec à Montréal, Montréal, QC, Canada7Centre d’Études Avancées en Médecine du Sommeil, Hôpital du Sacré-Coeur de Montréal,
Montréal, QC, Canada8Département de psychologie, Université de Montréal, Montréal, QC, Canada9Département de réadaptation, Université Laval, Québec, QC, Canada10Département de neurologie, Université McGill, Montréal, QC, Canada11Neurorehabilitation Research Centre, Montréal, QC, Canada12Centre de recherche, Centre Hospitalier de l'Université de Montréal, Montréal, QC, Canada
Corresponding author: Joël Macoir; Département de réadaptation, Faculté de médecine,
Pavillon Ferdinand-Vandry, Université Laval; 1036, rue de la Médecine, Bureau 4453; Québec,
QC, Canada, G1V 0A6; +1 418-656-2131 poste 12190; email: [email protected]
ASH: [email protected] CP: [email protected];
OP: [email protected]; LB: [email protected]; JFG: gagnon.jean-
[email protected]; SJ: [email protected]; SB: [email protected]; MW:
[email protected]; LK: [email protected]; IR: [email protected];
CH: [email protected]; JM: [email protected]
1
TMT: Normative Data for Quebec French-speaking Adults
Trail Making Test A and B: Regression-Based Normative data for Quebec
French-speaking Mid and Older-Aged Adults
Abstract
Objective: The Trail Making Test (TMT) is mainly used to assess visual scanning/processing
speed (part A) and executive functions (part B). The test has proven sensitive at detecting
cognitive impairment during aging. However, previous studies have shown differences between
normative data from different countries and cultures, even when corrected for age and education.
Such inconsistencies between normative data may lead to serious diagnostic errors, thus, the
development of local norms is warranted. The purpose of this study was to provide regression-
based normative data for TMT-A and -B, tailored for a large sample of French-speaking adults
from Quebec (Canada). Method: The normative sample consisted of 792 participants aged 50 to
91 years. Based on multiple linear regression, equations to calculate Z-scores were provided for
TMT-A and -B, and for a contrast score which compared performance between TMT-A and -B.
Percentiles, stratified by age, are presented for the number of recorded errors. Results: Age was
a significant predictor for TMT-A performance, while age and education were independently
associated with performance on TMT-B. Gender did not have any effect on performance, in
either condition. Education was the only significant predictor of the contrast score between
TMT-B and TMT-A. Examiners should remain vigilant when two or more errors are recorded on
the TMT-B since this was uncommon in the normative sample. Conclusions: Our TMT
normative data improve the accurate detection of visual scanning/processing speed and executive
function deficits in Quebec (Canada) French-speaking adults.
Key words: Norms/normative studies; Executive Functions; Attention; Trail Making Test;
Elderly/Geriatrics/Aging.
Word count: 4541 (including in the text authors’ names)
2
TMT: Normative Data for Quebec French-speaking Adults
Introduction
The Trail Making Test (TMT) is widely used in clinical settings to assess visual
scanning/psychomotor processing speed (Part A), and executive functions (Part B; primarily
cognitive flexibility/shifting) (Lezak, Howieson, Bigler, & Tranel, 2012; Salthouse, 2011). TMT
test performance is evaluated based on the speed of task completion as well as the number of
errors recorded during task execution. Both indices are important because they monitor different
cognitive processes (psychomotor speed vs. working memory/executive functions, respectively)
(Mahurin et al., 2006). Errors are uncommon in healthy participants, thus serving as sensitive
indicators of cerebral dysfunction (Mahurin et al., 2006). Many neuroimaging studies indicate
that TMT-B performance (especially the number of errors) is specifically related to cerebral
activity in the dorsolateral prefrontal region, because of its role in cognitive flexibility
(Davidson, Gao, Mason, Winocur, & Anderson, 2008; Moll, de Oliveira-Souza, Moll, Bramati,
& Andreiuolo, 2002; Stuss et al., 2001; Yochim, Baldo, Nelson, & Delis, 2007; Zakzanis, Mraz,
& Graham, 2005). However, a recent study showed that TMT-B performance did not differ
significantly in a group of patients with frontal brain lesions, compared to those with non-frontal
lesions. This suggests that the test is sensitive to general brain dysfunction, but not capable of
detecting damage in any specific brain region (Chan et al., 2015).
The TMT has also proved useful for the assessment of mild to severe traumatic brain
injury (Azouvi et al., 2016; de Guise et al., 2016), psychiatric conditions (Cotrena, Branco,
Shansis, & Fonseca, 2016), and neurodegenerative diseases (Ashendorf et al., 2008; Rasmusson,
Zonderman, Kawas, & Resnick, 1998; Roca et al., 2013). Recent studies have shown that TMT
performance was one of the best predictors of conversion from mild cognitive impairment to
dementia (Eckerstrom et al., 2015), especially dementia with Lewy bodies (Grenier Marchand,
Montplaisir, Postuma, Rahayel, & Gagnon, 2017). In a recent meta-analysis on the relationship
3
TMT: Normative Data for Quebec French-speaking Adults
between cognition and functional status, the TMT-B was shown to be a strong predictor of
everyday functioning in individuals with mild cognitive impairment (McAlister, Schmitter-
Edgecombe, & Lamb, 2016). The test is also associated with fitness to drive in older adults
(Bennett, Chekaluk, & Batchelor, 2016; Dickerson, Meuel, Ridenour, & Cooper, 2014).
Although task speed and the number of errors on the TMT-B more often identify pathology than
the TMT-A scores taken alone, authors argue that an algorithm that considers both speed and
errors are more sensitive markers of impairment (Ashendorf et al., 2008). Some scores seem to
be more useful, for example, time to completion of TMT-B (Eckerstrom et al., 2015; Grenier
Marchand et al., 2017) and total errors (Ashendorf et al., 2008), in predicting neurodegenerative
diseases. Time to completion of TMT-B and contrast scores between TMT-A and TMT-B have
also been shown to differentiate patients who will develop Parkinsonism from those who will
first develop dementia (Grenier Marchand et al., 2017).
In order to draw valid conclusions from an examinee’s performance, it is necessary to
employ normative data that control for the influence of sociodemographic variables (Mitrushina,
Boone, & D'Elia, 1999; Soukup, Ingram, Grady, & Schiess, 1998). A comparison of TMT
normative data from several countries and cultures has shown that norms across countries are
different, even when age, sex, and education are comparable across the samples (Fernández &
Marcopulos, 2008). Normative data are also different between Canada and the United States,
which have a similar western educational system (Mitrushina et al., 1999; Soukup et al., 1998).
Therefore, it is possible that using normative data derived from a different country or culture will
lead to unreliable results. Local norms are also more precise at detecting cognitive impairment
than non-cultural-specific norms (Arsenault-Lapierre et al., 2011).
A previous study that calculated normative data using Canadian English-speaking
participants suffers from a number of limitations (Tombaugh (2004). Tombaugh (2004)’s sample
4
TMT: Normative Data for Quebec French-speaking Adults
included participants who may have cognitive impairment or depression. Inclusion criteria
required a score of 23 or higher on the Mini-Mental State Examination (MMSE) (Folstein,
Folstein, & McHugh, 1975). A more conservative score of 26 or higher on the MMSE, or the
Montreal Cognitive Assessment test (MoCA), would aid exclusion of patients with mild
cognitive impairment (Nasreddine et al., 2005). Tombaugh (2004) also included participants with
a score of 14 and lower on the Geriatric Depression Scale (GDS), which is higher than the
suggested cut-off score of 11 (Yesavage, 1988). Additionally, the younger group included in the
normative data was exclusively composed of university students. This may limit the applicability
to populations with less education. The lowest performance from Tombaugh’s sample
corresponded to the 10th percentile (Z = -1.28), which does not meet the typical threshold for
statistical significance (alpha level of .05; 5th percentile; Z = -1.65). Moreover, very few studies
have calculated normative data for the number of errors recorded on the TMT (Amieva et al.,
2009; Ashendorf et al., 2008; Hankee et al., 2013).
Given the limitations mentioned above, the objective of the present study was to calculate
regression-based normative data for the TMT-A and -B, based on a large sample of Quebec
(Canada) French-speaking Mid and Older-Aged Adults. Studies have shown that norms based on
regression equations are useful to compare actual individual scores to predicted scores reflecting
specific demographic characteristics (Crawford & Howell, 1998). By using results from the
entire research sample, regression equations provide more stable and valid norms for any
subgroup. Regression equations have been used in several Quebec-French normative studies
(Callahan et al., 2014; Escudier et al., 2016; Larouche et al., 2016; Lavoie et al., 2018; St-Hilaire
et al., 2018; St-Hilaire et al., 2016; Tremblay et al., 2016; Tremblay et al., 2015).
Methods
5
TMT: Normative Data for Quebec French-speaking Adults
Participants
Researchers from across the province of Quebec (Canada) were invited to share anonymized data
from healthy French-speaking volunteers who had completed TMT-A and -B as part of other
research studies. All participants were mother tongue French (about 78% of the Quebec
population are mother tongue French (Government of Quebec, 2016)). All participants had
completed both TMT-A and -B conditions. All studies were approved by local Research Ethics
Boards and consent for the secondary use of these data was obtained during the primary studies.
The seven research sites which took part in this study were affiliated with universities in
Montreal and Quebec City, where participants were recruited and tested. The same instructions
were given to all participants according to a standardized protocol (Bowie & Harvey, 2006).
All participants scored within the normal range on the MMSE (≥ 26; Folstein et al., 1975)
or the MoCA (≥ 26; Nasreddine et al., 2005). In addition, participants had no clinically
significant depressive symptoms based on scores from the GDS (Yesavage, 1988), or the Beck
Depression Inventory second edition (BDI-II) (Beck, Steer, & Brown, 1996). Cut-offs for
inclusion were ≤ 10 for the 30-item GDS, ≤ 4 for the 15-item GDS, and ≤ 10 for the BDI-II. All
individuals self-reported good mental and physical health (i.e., no history of neurological
disease, untreated current psychiatric illness, traumatic brain injury, or any untreated medical
condition that could interfere with cognitive performance).
The normative sample consisted of 792 community-dwelling participants (510 women
and 282 men), aged between 50 and 91 years (mean age = 67.8 years; SD = 7.1) and having
between 3 and 23 years of formal education (mean education level = 14.9 years; SD = 3.5). The
majority of participants had received a high-school diploma (≥12 years) (83.7%; n = 663).
Women were somewhat overrepresented in our sample, at 64.4% (vs. 50.3% in overall Quebec
population) (Government of Canada, 2015).
6
TMT: Normative Data for Quebec French-speaking Adults
Materials and procedure
TMT paper sheets from Reitan (1979) were used for this study. No interference task between
TMT-A and -B conditions was performed. No time limit was set for task completion.
TMT-A is a paper sheet that contains circles, with the numbers 1 to 25, that are
distributed spatially in a semi-random order. The examinee is asked to draw a line connecting the
numbered circles, as fast as possible, in an ascending order. A practice trial with the numbers 1
to 8 is performed beforehand to ensure the participant understands the instructions. TMT-B is a
paper sheet that contains circles, with the numbers 1 to 13 and letters A to L, that are distributed
spatially in a semi-random order. The participant is instructed to draw a line connecting the
numbers and letters, as fast as possible, respecting an ascending and alphabetical order, by
alternating between numbers and letters (e.g., 1-A-2-B-3-C, etc.). A practice trial containing
eight items is performed beforehand to ensure the participant understands the instructions.
We followed the administration procedures and interpretive guidelines developed by
Bowie and Harvey (2006). The tests were administered in the French language. For both
conditions, mistakes were recorded and immediately corrected by the examiner, who drew an
“X” on the wrong connection. The participant was then instructed to return to the last properly
connected circle and continue the task from that point. The stopwatch would continue to run
during this time. Errors for TMT-B were of two types (Mahurin et al., 2006): (1) sequencing or
tracking (i.e., skipping a number or a letter on TMT-A or -B); and (2) set-loss or perseverative
(i.e., connecting two numbers or letters without alternation on TMT-B (e.g.: 1-A-2-B-C-3…or 1-
A-2-B-3-4-D…)).
Statistical Analyses
To identify the sociodemographic characteristics influencing task performance, a multiple linear
regression analysis was performed for each dependent variable using age, education, and gender
7
TMT: Normative Data for Quebec French-speaking Adults
as predictors. The distribution of both TMT-A and -B time scores were skewed. For that reason,
a logarithmic transformation (log10 (time)) was applied for both conditions (Tabachnick & Fidell,
2013).
Age and education were entered in the analyses as continuous variables, while gender
was coded 0 for women and 1 for men. Interactions between predictors were tested, (continuous
variables were centered), but none were significant. Therefore, they were not retained in the final
models.
Some patients may exhibit disproportionate performance on the TMT-B (difficulty in
cognitive flexibility/shifting, perseveration) relative to TMT-A (visual scanning/psychomotor
processing speed). In order to highlight significant differences between the two conditions, a
contrast score was computed. This analysis was based on the same procedure described by Delis,
Kaplan, and Kramer (2001). First, the uncorrected raw scores for TMT-A and -B (time) were
converted to scaled scores that are normally distributed and have a mean of 10 and a standard
deviation of 3. Second, for each participant, TMT-A scaled scores were subtracted from TMT-B
scaled scores. Finally, the scaled score differences were converted to a new distribution of scaled
scores. A contrast Z-score corrected for sociodemographic variables was then calculated using
linear regression. A Z-score under -1.65 (5th percentile) highlights significantly more difficulties
on the TMT-B, compared to the TMT-A condition. Meanwhile a Z-score higher than 1.65
indicates lower performance on the TMT-A condition. Visual and statistical analyses were
conducted to verify the underlying assumptions of the regression models (distribution and
residual normality, homogeneity of variance, linearity, multicollinearity, and outliers) using
common criteria (Tabachnick & Fidell, 2013). Semi-partial Pearson correlations (sr) were used
to describe the relative importance of the predictors. Due to the skewed distribution of error
scores, in both conditions, Spearman correlations were computed to determine the
8
TMT: Normative Data for Quebec French-speaking Adults
sociodemographic variables that were associated with error scores. Percentiles were generated
and stratified for each condition according to the significant correlations. All statistical analyses
were performed using SPSS software (version 21.0), with the alpha level set at .05.
Results
Table 1 shows the demographic characteristics of the participants in the normative sample. Table
2 shows the regression coefficients and intercepts for each measure of the TMT. The TMT-A
model accounted for 9.4% of the variance while only age (p < .001) was a significant predictor,
F(3, 788) = 27.4, p < .001 (education: p = .080, sex: p = .918). Semi-partial correlations revealed
that age (sr = .288) was by far the best predictor of TMT-A time, when compared to education
(sr = -.059) and gender (sr = -.004).
The TMT-B model accounted for 14.5% of the variance, F(3, 788) = 44.5, p < .001, while
both age (p < .001) and education (p < .001) were significant (sex: p = .559). Semi-partial
correlations revealed that age (sr = .295) was the best predictor of TMT-B time, followed by
education (sr = -.193), and gender (sr = -.019).
Finally, the model for the contrast score, between the TMT-B and -A conditions,
accounted for 2.3% of the variance, F(3, 788) = 6.1 p < .001, while only education (p < .001)
was significant (age: p = .637, gender: p = .825). Semi-partial correlations revealed that
education (sr = .151) was a better predictor of the contrast score than age (sr = .017), or gender
(sr = .008).
Table 3 shows the Z-scores for TMT-A and -B based on the results from the regression
models. In order to facilitate the calculation of Z-scores based on the regression formulas, we
prepared a Microsoft Excel® spreadsheet containing automatic formulas. This file can be
9
TMT: Normative Data for Quebec French-speaking Adults
downloaded from the journal’s website (see Supplemental online material) or by writing to the
corresponding author.
The number of TMT-A and -B errors was available for 80.9% of the sample (n = 641).
No sociodemographic predictors were significantly correlated with the number of errors on the
TMT-A (age: rs = -.005, p = .906; education: rs = .009, p = .823; gender: rs = .028, p = .487). The
majority of participants did not commit any errors on the TMT-A (90.8%, n = 581). Only age
was correlated with the number of errors on the TMT-B (age: rs = .089, p = .025; education: rs =
-.074, p = .061; gender: rs = .002, p = .951). The majority of participants did not perform any
errors on the TMT-B (73.8%; n = 473). Only 45 participants (0.07%) made two or more errors
on the TMT-B. These participants were also slower on the TMT-B (p = .009) and performed
more errors on the TMT-A (p = .042) than the rest of the sample. However, these participants did
not differ from the whole sample with respect to age, education, gender, global cognition,
depression or TMT-A time. Percentiles for total error scores on TMT-A and -B are shown in
Table 4. Due to the low number of errors on the TMT-B, we did not attempt to distinguish
between the types of errors (i.e., sequencing or set-loss). Of note, the slowest participants also
made the greatest number of errors in both conditions; TMT-A (rs = .079, p = .045), and
especially TMT-B (rs = .347, p < .001).
Discussion
Previous studies have found that normative TMT data from different countries and cultures are
different (Fernández & Marcopulos, 2008). That is why the main objective of this study was to
establish normative data for the TMT-A and -B, based on a large sample of French-speaking
adults from Quebec, who were aged between 50 and 91 years. Linear multiple regressions were
performed for each condition using age, education, and gender as predictors. Results indicated
10
TMT: Normative Data for Quebec French-speaking Adults
that age and education were independently associated with performance on the TMT-B, while
only age was a significant predictor of TMT-A scores. Faster task speeds were apparent in
individuals with higher levels of education (TMT-B) and those who were younger (TMT-A and -
B), which is consistent with studies showing declines in processing speed during aging
(Salthouse, 1996). Correlations between sociodemographic variables and scores on the TMT-A
and -B showed that age was more highly associated with performance than was education. This
result echoes those of other normative studies conducted in North America (Ashendorf et al.,
2008; Ivnik, Malec, Smith, Tangalos, & Petersen, 1996; Tombaugh, 2004). Education may
account for more variance than age on TMT-B scores in normative data that use samples with
significantly older and less educated participants (Lucas et al., 2005). On the other hand,
Tombaugh (2004) found that very little variance was explained by education when only adults
aged 55 and older were included in the analysis (TMT-A (1.5%) and -B (4.4%)). Thus, it appears
that controlling for the effect of age is sufficient to interpret performance on the TMT-A and that
education explains slightly more of the performance on the TMT-B.
The present study found no effect of gender on TMT performance in either condition,
which has been shown in previous studies (Ashendorf et al., 2008; Lucas et al., 2005; Schneider
et al., 2015; Strauss, Sherman, & Spreen, 2006; Tombaugh, 2004). The only study that reported
an effect of gender was conducted in France by Amieva et al. (2009). However, despite being
statistically significant, we do not know whether this result is clinically significant.
Education was the only significant predictor of the contrast score between TMT-B and
TMT-A scaled scores. The contrast score allows us to determine whether cognitive
flexibility/shifting (TMT-B) is significantly deficient in comparison to psychomotor processing
speed (TMT-A). It is still possible to perform worse on the TMT-A, in comparison to TMT-B,
for many reasons (time required to become familiar with the task, lack of motivation, less effort
11
TMT: Normative Data for Quebec French-speaking Adults
due to the apparent ease of TMT-A, etc.). One should also note that a contrast score could be
considered ‘normal’ if both conditions were impaired. Thus, contrast scores must be interpreted
with caution. A Z-score below -1.65 (5th percentile) indicates significantly worse performance on
the TMT-B while a Z-score above 1.65 instead indicates significantly worse performance on the
TMT-A.
Few North American studies have proposed normative data for the number of TMT
errors. Yet, recorded errors are useful for distinguishing between psychiatric conditions
(Mahurin et al., 2006), the detection of malingering (Ruffolo, Guilmette, & Willis, 2000), and for
tracking progression from normal aging to mild cognitive impairment and dementia (Ashendorf
et al., 2008). Moreover, both TMT time and errors are useful because, while they are both
influenced by working memory and executive functions, error rates are not influenced by faster
psychomotor speed (Mahurin et al., 2006). On the contrary, we found that it is the slowest
participants who make the most mistakes. However, one cannot completely rule out the possible
influence of psychomotor speed on the error rate. Given that the time required for the participant
to correct an error is included in the TMT time to completion, it is indeed possible that for some
participants, faster psychomotor speed led to more errors and subsequently increased their time
on task. The examiners must be vigilant when one or more errors are performed on the TMT-A
because it is unusual (9.2%). Likewise, two or more errors on the TMT-B were very uncommon
(0.07%) and lead to a significant deficit in performance (≤ 5th percentile), irrespective of the
participant’s age, which is in line with findings from Stuss et al. (2001). Thus, it seems that the
number of errors does not increase dramatically during normal aging and is more an indicator of
impairment (Ashendorf et al., 2008). Consistent with the results of the current study, Ashendorf
et al. (2008) found no significant association between sociodemographic variables and the
number of errors on the TMT-A. In the present study, TMT-B error rates was only associated
12
TMT: Normative Data for Quebec French-speaking Adults
with age, while Ashendorf et al. (2008) found a significant association with both age and
education. Differences may arise from the fact that their sample was composed of more
individuals with lower education than in our sample.
Strengths and Limitations
Despite all labs administering the TMT according to a standardized protocol, it was not
possible to perform quality control of the data collection. Moreover, the current study used an
incidental sampling method; however, the normative data presented here were acquired from a
large sample of adults living in the two most populous areas in the Province of Quebec (Montreal
and Quebec City). Participant demographics covered a respectable range of ages and education
levels. However, greater score variability may be present in subsamples comprised of very old
adults (only 6.2% of the sample was aged 80-91 years) and individuals with low educational
levels (< 7 years), since there were relatively few cases in our sample. Results should be
interpreted with caution for these groups. In addition, the current normative data should not be
applied to individuals outside the stated age range since it would represent estimated scores.
Finally, the current sample contained more women than men. However, since gender did not
have a significant effect on task performance, the present results appear generalizable. While
some authors showed that ethnicity may have an impact on both conditions of TMT performance
(Schneider et al., 2015), this information was not available in our data. The majority of
individuals in Montreal and Quebec City are white Caucasians.
Standardized administration commonly allows 180 seconds to complete the TMT-A
(Strauss et al., 2006) and 300 seconds to complete the TMT-B (Heaton, Miller, Taylor, & Grant,
2004). There was no time-limit in our study because it was found that this practice may mask
performance variability, especially among cognitively impaired individuals who cannot complete
the task (Correia et al., 2015). However, none of the participants in our sample had a completion
13
TMT: Normative Data for Quebec French-speaking Adults
time of over 180 seconds for TMT-A (range: 15 to 102), and only three participants had a
completion time of over 300 seconds for TMT-B (range: 25 to 348).
An original aspect of this study was the regression-based approach used to calculate
normative data for completion times on the TMT-A and -B. First, this normative method has the
advantage of better estimating the expected performance of a participant given their personal
characteristics, instead of discrete norms created using arbitrary age groups. In the latter case, the
relative standing of an individual may change dramatically as they move from one age category
to another (Crawford & Howell, 1998). Second, by contrast to norms based on means or
percentiles of arbitrary sociodemographic subgroups, norms based on regression equations have
shown to reduce demographic biases associated with the use of raw data in neuropsychological
tests (e.g., Heaton, Avitable, Grant, and Matthews (1999), Van der Elst, Van Boxtel, Van
Breukelen, and Jolles (2006)). Third, in addition to identifying which variables are relevant to
the norming, the regression-based approach also provides more stable norms for any subgroup by
using data from the entire participant sample (Van Breukelen & Vlaeyen, 2005)
To illustrate the advantage of regression equations, let us imagine a 75-year-old-man with
12 years of education who completed TMT-A in 66 seconds and TMT-B in 178 seconds. First,
based on the regression equations from Table 3, the patient’s Z-scores would be -1.32 (9th
percentile) and -1.55 (6th percentile), for TMT-A and TMT-B respectively. These results appear
to be indicative of slight to moderate slowness in processing speed and/or difficulty with
cognitive flexibility. If the results of this participant were instead compared to the Canadian
normative data from Tombaugh (2004), his Z-scores would be -1.28 (10th percentile), for both
TMT conditions. These results would be considered weak but would remain normal. Now, let us
rather imagine that this man who completed TMT-A in 66 seconds and TMT-B in 178 seconds:
(a) is aged 75 and has 7 years of education or (b) is aged 55 and has 12 years of education. Based
14
TMT: Normative Data for Quebec French-speaking Adults
on the regression equations, the Z-scores for the two TMT conditions would be -1.24 and -1.25
(11th percentile) for the hypothetical case (a) and -2.24 and -2.47 (< 1st percentile) for the
hypothetical case (b), for TMT-A and TMT-B respectively. Results for case (a) indicate slight
slowness while it is an indicator of deficit for the youngest man (b). Using normative data from
Tombaugh (2004), case (a) obtains higher scores (15th percentile; Z = -1.00) while we know at
best that patient (b) has a score below the 10th percentile (the smallest percentile for these
normative data), without knowing exactly which one. We remind the reader that Tombaugh
(2004) sample included individuals with MMSE scores ≥ 24 (instead of ≥ 26) and individuals
with GDS-30 scores ≤ 13 (instead of ≤ 10). This increases the risk of including participants with
mild cognitive impairment (Nasreddine et al., 2005) or active depressive symptomatology in the
normative data (Yesavage, 1988) and possibly lengthening the task completion time. Therefore,
our study is more likely to only include participants who do not have any cognitive impairment
or depression. With regards to cultural differences (e.g., Quebec French vs European French), we
compared the first hypothetical patient’s Z-scores (i.e. -1.32 and -1.55 for TMT-A and TMT-B
respectively) to those obtained from Amieva et al. (2009) European French normative data. The
differences are more impressive. Indeed, using Amieva et al. (2009) normative data, this 75-
year-old-man would reach a normal performance for TMT-A (Z = -0.67; 25th percentile) and a
somewhat inaccurate score on TMT-B (Z between -1.65 and -1.28; between 5th and 10th
percentile). Finally, compared to the most widely used normative data in the United States,
Heaton et al. (2004) (ZTMTA = -1.60; 5th percentile and ZTMTB = -1.00; 16th percentile) and
Mitrushina et al. (1999) (ZTMTA and –B = -1.00; 16th percentile), our normative data were once again
more sensitive at detecting deficits in psychomotor processing speed or executive functions, at
least on TMT-B. As stated by Fernández and Marcopulos (2008), these various examples
underline the importance of using culturally derived normative data. However, further studies are
15
TMT: Normative Data for Quebec French-speaking Adults
needed to establish the sensibility and specificity of our normative data in detecting cognitive
impairment in clinical populations. Our norms should only be used for Quebec French-speaking
adults. With regards to the use of the current norms with French-speaking Canadians living
outside Quebec, this practice should be applied with caution since the data of the present study
only come from French Quebecers. In sum, data from the present study will strengthen accurate
detection of deficits in psychomotor processing speed and executive functions in Quebec French-
speaking adults and will invite clinicians to push further their investigation of potential executive
deficits with additional testing.
Acknowledgements
CH, SJ, and JFG were supported by a salary award from the FRQ-S. JFG holds a Canada Research Chair
in Cognitive Decline in Pathological Aging and a salary award from the CIHR.We are grateful to Scott
Nugent and Lynn Maynard for their help with the English revision of this article.
Disclosure Statement
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or
publication of this article.
Funding
This work was supported by the Réseau québécois de recherche sur le vieillissement, the Canadian
Institutes of Health Research (CIHR), the Alzheimer Society of Canada, the Natural Sciences and
Engineering Research Council of Canada, the Fonds de Recherche du Québec – Santé (FRQ-S), and the
Fonds de recherche du Québec – Société et culture (FRQSC).
16
TMT: Normative Data for Quebec French-speaking Adults
Table 1. Demographics of participants in the normative sample (n = 792)
17
Characteristics N %
Age 50-59 80 10,1 60-64 177 22.3 65-69 234 29.5 70-74 170 21.5 75-79 82 10.4 80-91 49 6.2
Gender (men/women) 282/510 35.6/64.4
Education Elementary (3-7 years) 15 1.9 High-School (8-12 years) 195 24.6 College (13-14 years) 140 17.7 University undergraduate (15-17 years) 238 30.1 University graduate (18-19 years) 132 16.7 University postgraduate (20-23 years) 72 9.1
TMT: Normative Data for Quebec French-speaking Adults
Table 2. Coefficients for multiple linear regression analyses for TMT measures
Variable B β t p
TMT-Aa
Age 0.006 0.292 8.49 <.001Education -0.002 -0.061 -1.75 .080Sex -0.001 -0.004 -0.10 .918
TMT-Bb
Age 0.007 0.299 8.95 <.001Education -0.009 -0.197 -5.87 <.001Sex -0.007 -0.019 -0.59 .559
Contrast scorec
Age 0.007 0.017 0.47 .637Education 0.131 0.153 4.28 <.001Sex 0.049 0.008 0.22 .825
a Prediction for the log10 TMT-A time. Intercept = 1.221; Square root of the mean square residual = 0.130.b Prediction for the log10 TMT-B time. Intercept = 1.592; Square root of the mean square residual = 0.152.c Intercept = -12.475; Square root of the mean square residual = 2.971.
18
TMT: Normative Data for Quebec French-speaking Adults
Table 3. Normative equations to calculate corrected Z-scores for age, education and gender for TMT-A and -B (n = 792)
Variable Corrected Z-score for age, education and gender
TMT-A Z score (log10(TMT-A time) - (1.221 + (0.006*Age) + (-0.002*Education) + (0.001*Gender))) / 0.1304 * -1
TMT-B Z score (log10(TMT-B time) - (1.592 + (0.007*Age) + (-0.009*Education) + (0.007*Gender))) / 0.1517 * -1
Contrast score
TMT-A SS 3*((TMT-A time - 38.359) / 12.836) + 10
TMT-B SS 3*((TMT-B time - 88.014369) / 39.157) + 10
Contrast SS (3*(((TMT-B SS - TMT-A SS) - -8.301E-8) / 2.729) + 10) * -1
Contrast Z score (Contrast scale score - (-12.475 + (0.007*Age) + (0.131*Education) + (-0.049*Gender))) / 2.9712
Notes: Age: participant’s age (continuous variable; between 50 and 91); Education: years of education (continuous variable; between 3 and 23). Gender: female=0, male=1. Equation denominators corresponded to residual standard deviations of each model. Multiplication by -1 makes it possible to obtain a negative Z-score for slower performance. SS: scale score.
19
TMT: Normative Data for Quebec French-speaking Adults
Table 4. Percentiles for error scores (n = 641)
20
Measure nPercentiles for the number of errors
1st 2nd 5th 10th 15th 25th 50th 95th
TMT-A 641 2 1 1 0 0 0 0 0
TMT-B
Age 50-69 413 3 3 2 1 1 0 0 0
Age 70-91 228 4 3 2 1 1 1 0 0
TMT: Normative Data for Quebec French-speaking Adults
ReferencesAmieva, H., Le Goff, M., Stoykova, R., Lafont, S., Ritchie, K., Tzourio, C., . . . Dartigues, J.-F.
(2009). Trail Making Test A et B (version sans correction des erreurs): Normes en population chez des sujets âgés, issues de l'étude des trois Cités. [Trail Making Test: Normative data in a French elderly population from the three Cities Study.]. Revue de Neuropsychologie, Neurosciences Cognitives et Cliniques, 1, 210-220. doi: 10.3917/rne.013.0210
Arsenault-Lapierre, G., Whitehead, V., Belleville, S., Massoud, F., Bergman, H., & Chertkow, H. (2011). Mild cognitive impairment subcategories depend on the source of norms. Journal of Clinical and Experimental Neuropsychology, 33, 596-603. doi: 10.1080/13803395.2010.547459
Ashendorf, L., Jefferson, A. L., O'Connor, M. K., Chaisson, C., Green, R. C., & Stern, R. A. (2008). Trail Making Test errors in normal aging, mild cognitive impairment, and dementia. Archives of Clinical Neuropsychology, 23, 129-137. doi: 10.1016/j.acn.2007.11.005
Azouvi, P., Vallat-Azouvi, C., Joseph, P. A., Meulemans, T., Bertola, C., Le Gall, D., . . . Godefroy, O. (2016). Executive Functions Deficits After Severe Traumatic Brain Injury: The GREFEX Study. Journal of Head Trauma Rehabilitation, 31, E10-20. doi: 10.1097/htr.0000000000000169
Barr, W. B. (2003). Neuropsychological testing of high school athletes: Preliminary norms and test-retest indices. Archives of Clinical Neuropsychology, 18, 91-101. doi: 10.1016/S0887-6177(01)00185-8
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Beck Depression Inventory-II: Manual. San Antonio, TX: The Psychological Corporation.
Bennett, J. M., Chekaluk, E., & Batchelor, J. (2016). Cognitive Tests and Determining Fitness to Drive in Dementia: A Systematic Review. Journal of the American Geriatrics Society, 64, 1904-1917. doi: 10.1111/jgs.14180
Bowie, C. R., & Harvey, P. D. (2006). Administration and interpretation of the Trail Making Test. Nature Protocols, 1, 2277-2281. doi: 10.1038/nprot.2006.390
Callahan, B. L., Belleville, S., Ferland, G., Potvin, O., Tremblay, M. P., Hudon, C., & Macoir, J. (2014). Normative data for a computer-assisted version of the auditory three-consonant Brown-Peterson paradigm in the elderly French-Quebec population. The Clinical Neuropsychologist, 28, 317-332. doi: 10.1080/13854046.2013.873082
Chan, E., MacPherson, S. E., Robinson, G., Turner, M., Lecce, F., Shallice, T., & Cipolotti, L. (2015). Limitations of the trail making test part-B in assessing frontal executive dysfunction. Journal of the International Neuropsychological Society, 21, 169-174. doi: 10.1017/s135561771500003x
Correia, S., Ahern, D. C., Rabinowitz, A. R., Farrer, T. J., Watts, A. K. S., Salloway, S., . . . Deoni, S. C. L. (2015). Lowering the floor on trail making test part B: Psychometric evidence for a new scoring metric. Archives of Clinical Neuropsychology, 30, 643-656. doi: 10.1093/arclin/acv040
Cotrena, C., Branco, L. D., Shansis, F. M., & Fonseca, R. P. (2016). Executive function impairments in depression and bipolar disorder: association with functional impairment and quality of life. Journal of Affective Disorders, 190, 744-753. doi: 10.1016/j.jad.2015.11.007
Crawford, J. R., & Howell, D. C. (1998). Regression equations in clinical neuropsychology: an evaluation of statistical methods for comparing predicted and obtained scores. Journal of
21
TMT: Normative Data for Quebec French-speaking Adults
Clinical and Experimental Neuropsychology, 20, 755-762. doi: 10.1076/jcen.20.5.755.1132
Davidson, P. S., Gao, F. Q., Mason, W. P., Winocur, G., & Anderson, N. D. (2008). Verbal fluency, trail making, and Wisconsin Card Sorting Test performance following right frontal lobe tumor resection. Journal of Clinical and Experimental Neuropsychology, 30, 18-32. doi: 10.1080/13803390601161166
de Guise, E., Belanger, S., Tinawi, S., Anderson, K., LeBlanc, J., Lamoureux, J., . . . Feyz, M. (2016). Usefulness of the rivermead postconcussion symptoms questionnaire and the trail-making test for outcome prediction in patients with mild traumatic brain injury. Applied Neuropsychology. Adult., 23, 213-222. doi: 10.1080/23279095.2015.1038747
Delis, D. C., Kaplan, E., & Kramer, J. H. (2001). Delis-Kaplan Executive Function System (D–KEFS) San Antonio: TX: The Psychological Corporation.
Dickerson, A. E., Meuel, D. B., Ridenour, C. D., & Cooper, K. (2014). Assessment tools predicting fitness to drive in older adults: a systematic review. American Journal of Occupational Therapy, 68, 670-680. doi: 10.5014/ajot.2014.011833
Eckerstrom, C., Olsson, E., Klasson, N., Berge, J., Nordlund, A., Bjerke, M., & Wallin, A. (2015). Multimodal prediction of dementia with up to 10 years follow up: the Gothenburg MCI study. Journal of Alzheimer's Disease, 44, 205-214. doi: 10.3233/jad-141053
Escudier, F., Leveille, E., Charbonneau, S., Cole, J., Hudon, C., Bedirian, V., & Scherzer, P. (2016). Evaluating Decision-Making: Validation and Regression-Based Normative Data of the Judgment Assessment Tool. Archives of Clinical Neuropsychology. doi: 10.1093/arclin/acw019
Fernández, A. L., & Marcopulos, B. A. (2008). A comparison of normative data for the Trail Making Test from several countries: Equivalence of norms and considerations for interpretation. Scandinavian Journal of Psychology, 49, 239-246. doi: 10.1111/j.1467-9450.2008.00637.x
Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). "Mini-mental state". A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, 189-198. doi: 10.1016/0022-3956(75)90026-6
Government of Canada. (2015). Population by sex and age group, by province and territory (Proportion of, female). Statistics Canada, Retrieved from http://www.statcan.gc.ca/tables-tableaux/sum-som/l02/cst01/demo31f-fra.htm.st.
Government of Quebec. (2016). Québec Handy Numbers, 2016 Edition. Institut de la statistique du Québec, Retrieved from http://www.stat.gouv.qc.ca/quebec-chiffre-main/pdf/qcm2016_an.pdf.
Grenier Marchand, D., Montplaisir, J., Postuma, R. B., Rahayel, S., & Gagnon, J. F. (2017). Detecting the Cognitive Prodrome of Dementia with Lewy Bodies: A Prospective Study of REM Sleep Behavior Disorder. Sleep, 40. doi: 10.1093/sleep/zsw014
Hankee, L. D., Preis, S. R., Beiser, A. S., Devine, S. A., Liu, Y., Seshadri, S., . . . Au, R. (2013). Qualitative neuropsychological measures: Normative data on executive functioning tests from the Framingham Offspring Study. Experimental Aging Research, 39, 515-535. doi: 10.1080/0361073X.2013.839029
Heaton, R. K., Avitable, N., Grant, I., & Matthews, C. G. (1999). Further crossvalidation of regression-based neuropsychological norms with an update for the Boston Naming Test. Journal of Clinical and Experimental Neuropsychology, 21, 572-582. doi: 10.1076/jcen.21.4.572.882
22
TMT: Normative Data for Quebec French-speaking Adults
Heaton, R. K., Miller, S. W., Taylor, M. J., & Grant, I. (2004). Revised comprehensive norms for an expanded Halstead Reitan battery: Demographically adjusted neuropyschological norms for African Americans and Caucasian adults. Lutz, FL: Psychological Assessment Resources.
Ivnik, R. J., Malec, J. F., Smith, G. E., Tangalos, E. G., & Petersen, R. C. (1996). Neuropsychological tests' norms above age 55: COWAT, BNT, MAE Token, WRAT-R Reading, AMNART, STROOP, TMT, and JLO. The Clinical Neuropsychologist, 10, 262-278. doi: 10.1080/13854049608406689
Ivnik, R. J., Malec, J. F., Smith, G. E., Tangalos, E. G., & Petersen, R. C. (1996). Neuropsychological tests' norms above age 55: COWAT, BNT, MAE Token, WRAT-R Reading, AMNART, STROOP, TMT, and JLO. Clinical Neuropsychologist, 10, 262-278. doi: 10.1080/13854049608406689
Larouche, E., Tremblay, M. P., Potvin, O., Laforest, S., Bergeron, D., Laforce, R., . . . Hudon, C. (2016). Normative Data for the Montreal Cognitive Assessment in Middle-Aged and Elderly Quebec-French People. Archives of Clinical Neuropsychology. doi: 10.1093/arclin/acw076
Lavoie, M., Bherer, L., Joubert, S., Gagnon, J. F., Blanchet, S., Rouleau, I., . . . Hudon, C. (2018). Normative data for the Rey Auditory Verbal Learning Test in the older French-Quebec population. The Clinical Neuropsychologist. doi: 10.1080/13854046.2018.1429670
Lezak, M. D., Howieson, D. B., Bigler, E. D., & Tranel, D. (2012). Neuropsychological assessment (5th ed.). New York, NY, US: Oxford University Press.
Lucas, J. A., Ivnik, R. J., Smith, G. E., Ferman, T. J., Willis, F. B., Petersen, R. C., & Graff-Radford, N. R. (2005). Mayo's Older African Americans Normative Studies: Norms for Boston Naming Test, Controlled Oral Word Association, Category Fluency, Animal Naming, Token Test, WRAT-3 Reading, Trail Making Test, Stroop Test, and Judgment of Line Orientation. The Clinical Neuropsychologist, 19, 243-269. doi: 10.1080/13854040590945337
Mahurin, R. K., Velligan, D. I., Hazleton, B., Davis, J. M., Eckert, S., & Miller, A. L. (2006). Trail making test errors and executive function in schizophrenia and depression. The Clinical Neuropsychologist, 20, 271-288. doi: 10.1080/13854040590947498
McAlister, C., Schmitter-Edgecombe, M., & Lamb, R. (2016). Examination of Variables That May Affect the Relationship Between Cognition and Functional Status in Individuals with Mild Cognitive Impairment: A Meta-Analysis. Archives of Clinical Neuropsychology, 31, 123-147. doi: 10.1093/arclin/acv089
Mitrushina, M. N., Boone, K. B., & D'Elia, L. F. (1999). Handbook of normative data for neuropsychological assessment. New York, NY, US: Oxford University Press.
Moll, J., de Oliveira-Souza, R., Moll, F. T., Bramati, I. E., & Andreiuolo, P. A. (2002). The cerebral correlates of set-shifting: an fMRI study of the trail making test. Arquivos de Neuro-Psiquiatria, 60, 900-905. doi: 10.1590/S0004-282X2002000600002
Nasreddine, Z. S., Phillips, N. A., Bédirian, V., Charbonneau, S., Whitehead, V., Collin, I., . . . Chertkow, H. (2005). The Montreal Cognitive Assessment, MoCA: A Brief Screening Tool For Mild Cognitive Impairment. Journal of the American Geriatrics Society, 53, 695-699. doi: 10.1111/j.1532-5415.2005.53221.x
Rasmusson, X. D., Zonderman, A. B., Kawas, C., & Resnick, S. M. (1998). Effects of Age and Dementia on the Trail Making Test. The Clinical Neuropsychologist, 12, 169-178. doi: 10.1076/clin.12.2.169.2005
23
TMT: Normative Data for Quebec French-speaking Adults
Reitan, R. M. (1979). Manual for administration of neuropsychological test batteries for adults and children. . Tucson, AZ: Reitan Neuropsychological Laboratories Inc,.
Roca, M., Manes, F., Gleichgerrcht, E., Watson, P., Ibáñez, A., Thompson, R., . . . Duncan, J. (2013). Intelligence and executive functions in frontotemporal dementia. Neuropsychologia, 51, 725-730. doi: 10.1016/j.neuropsychologia.2013.01.008
Ruffolo, L. F., Guilmette, T. J., & Willis, W. G. (2000). Comparison of time and error rates on the Trail Making Test among patients with head injuries, experimental malingerers, patients with suspect effort on testing, and normal controls. The Clinical Neuropsychologist, 14, 223-230. doi: 10.1076/1385-4046(200005)14:2;1-Z;FT223
Salthouse, T. A. (1996). The processing-speed theory of adult age differences in cognition. Psychological Review, 103, 403-428. doi: 10.1037/0033-295X.103.3.403
Salthouse, T. A. (2011). What cognitive abilities are involved in trail-making performance? Intelligence, 39, 222-232. doi: 10.1016/j.intell.2011.03.001
Schneider, A. L., Sharrett, A. R., Gottesman, R. F., Coresh, J., Coker, L., Wruck, L., . . . Mosley, T. H. (2015). Normative data for 8 neuropsychological tests in older blacks and whites from the atherosclerosis risk in communities (ARIC) study. Alzheimer Disease & Associated Disorders, 29, 32-44. doi: 10.1097/wad.0000000000000042
Soukup, V. M., Ingram, F., Grady, J. J., & Schiess, M. C. (1998). Trail Making Test: issues in normative data selection. Applied Neuropsychology, 5(2), 65-73. doi: 10.1207/s15324826an0502_2
St-Hilaire, A., Blackburn, M. C., Wilson, M. A., Laforce, R., Jr., Hudon, C., & Macoir, J. (2018). Object decision test (BORB): normative data for the adult Quebec population and performance in Alzheimer's disease and the semantic variant of primary progressive aphasia. Neuropsychology, Development, and Cognition. Section B, Aging, Neuropsychology and Cognition, 25, 427-442. doi: 10.1080/13825585.2017.1319901
St-Hilaire, A., Hudon, C., Vallet, G. T., Bherer, L., Lussier, M., Gagnon, J. F., . . . Macoir, J. (2016). Normative data for phonemic and semantic verbal fluency test in the adult French-Quebec population and validation study in Alzheimer's disease and depression. The Clinical Neuropsychologist, 30, 1126-1150. doi: 10.1080/13854046.2016.1195014
Strauss, E., Sherman, E. M. S., & Spreen, O. (2006). A compendium of neuropsychological tests: Administration, norms, and commentary (3rd. ed). New York, NY, US: Oxford University Press.
Stuss, D. T., Bisschop, S. M., Alexander, M. P., Levine, B., Katz, D., & Izukawa, D. (2001). The Trail Making Test: a study in focal lesion patients. Psychological Assessment, 13, 230-239. doi: 10.1037/1040-3590.13.2.230
Tabachnick, B. G., & Fidell, L. S. (2013). Using multivariate statistics (6th ed ed.). Boston: Pearson Education.
Tombaugh, T. N. (2004). Trail Making Test A and B: normative data stratified by age and education. Archives of Clinical Neuropsychology, 19, 203-214. doi: 10.1016/s0887-6177(03)00039-8
Tremblay, M. P., Potvin, O., Belleville, S., Bier, N., Gagnon, L., Blanchet, S., . . . Hudon, C. (2016). The Victoria Stroop Test: Normative Data in Quebec-French Adults and Elderly. Archives of Clinical Neuropsychology. doi: 10.1093/arclin/acw029
Tremblay, M. P., Potvin, O., Callahan, B. L., Belleville, S., Gagnon, J. F., Caza, N., . . . Macoir, J. (2015). Normative data for the Rey-Osterrieth and the Taylor complex figure tests in Quebec-French people. Archives of Clinical Neuropsychology, 30, 78-87. doi: 10.1093/arclin/acu069
24
TMT: Normative Data for Quebec French-speaking Adults
Van Breukelen, G. J. P., & Vlaeyen, J. W. S. (2005). Norming clinical questionnaires with multiple regression: The Pain Cognition List. Psychological Assessment, 17, 336-344. doi: 10.1037/1040-3590.17.3.336
Van der Elst, W., Van Boxtel, M. P., Van Breukelen, G. J., & Jolles, J. (2006). The Stroop color-word test: influence of age, sex, and education; and normative data for a large sample across the adult age range. Assessment, 13, 62-79. doi: 10.1177/1073191105283427
Yesavage, J. A. (1988). Geriatric Depression Scale. Psychopharmacology Bulletin, 24, 709-711. Yochim, B., Baldo, J., Nelson, A., & Delis, D. C. (2007). D-KEFS Trail Making Test
performance in patients with lateral prefrontal cortex lesions. Journal of the International Neuropsychological Society, 13, 704-709. doi: 10.1017/s1355617707070907
Zakzanis, K. K., Mraz, R., & Graham, S. J. (2005). An fMRI study of the Trail Making Test. Neuropsychologia, 43, 1878-1886. doi: 10.1016/j.neuropsychologia.2005.03.013
25